There are obviously a number of people contributing to this thread who have a limited practical experience of operating for sustained periods in moderate icing and who additionally have only a cursory knowledge of a/c anti/de-ice systems.
An a/c does not simply have a particular de-ice or anti-ice system that is switched on/off with a consistant result rather like a light bulb. Crew have to be aware of current icing conditions and when possible help alleviate the problem by altitude or route changes. In temperate latitudes ice accretion can be markedly reduced simply by climbing or descending 3000 ft. One crew may make a decision to climb or descend in cruise, the crew of a similar a/c following along behind may simply stay at their planned cruise altitude and experience problems that the first crew avoided. One crew having avoided icing by climbing may request a late descent to destination so as to avoid prolonged exposure to ice in the descent... another crew might not bother.
As for extending the boots through an auto-system that doesn't wait for any ice to build or making a decision to wait for a build up of 1/2 to one inch of ice.. this decision should be made in accordance with the a/c manufacturers advice... a/c boots operate differently on different a/c... the manufacturer carries out the tests, follow their advice.
Again, one crew may decide to continue a flight utilising the auto-pilot and another crew, correctly disengaging it and flying manually... auto-pilots disguise the onset of trim changes required to keep the a/c flying that would otherwise indicate to a crew that things are getting seriously out of hand. On disengaging the autopilot the crew can be in for a rude awakening.
Tail icing (as I posted earlier) needs to be recognised as such and the correct recovery technique applied... with tail icing column buffet will be experienced (as opposed to feeling the buffet through the airframe as with a conventional wing stall)... the column will have a tendency to move easily foward and be increasingly difficult to move aft giving rise to Pilot Induced Oscillations... if allowed to develop the column will move sharpley forward as a full stall develops... the recovery procedure is to pull back on the column (opposite of a conventional wing stall) though by this time the forces required may be so great as to prevent the crew from recovering, especially as this event is likely to occur during a configuration change at a lower level during the approach.
This crew appears to have followed the recommended procedure in 'undoing' the configeration changes that may have caused the initial upset by selecting flaps to their original position and gear up... all sadly too late to recover. The problem with icing accidents is that all too often the evidence is lost in the resulting impact/fire.
Hopefully the FDR will offer some clues... in the meantime let's all keep in mind that, as with wind shear, ice is something best (when possible) avoided.